Viral Drug Resistance
The use of anti-viral compounds for chemotherapy and chemoprophylaxis of viral diseases is a relatively new development in the field of infectious diseases, particularly when compared with the more than 50 years of experience with antibacterial antibiotics. The design of anti-viral compounds is not straightforward because viruses present a number of unique problems. Viruses must replicate intracellularly and often employ host cell enzymes, macromolecules, and organelles for the synthesis of virus particles. Therefore, safe and effective anti-viral compounds must be able to discriminate with a high degree of efficiency between cellular and virus-specific functions. In addition, because of the nature of virus replication, evaluation of the in vitro sensitivity of virus isolates to anti-viral compounds must be carried out in a complex culture system consisting of living cells (e.g. tissue culture). The results from such assay systems vary widely according to the type of tissue culture cells which are employed and the conditions of assay. Despite these complexities nine drugs have been approved for AIDS therapy, five reverse transcriptase inhibitors AZT, ddI, ddC, d4T, 3TC, one non-nucleoside reverse transcriptase inhibitor, nevirapine and three protease inhibitors saquinavir, ritonavir and indinovir and several additional anti-viral drug candidates have been recently developed such as nelfinavir, delaviridine, VX-478 and 1592.
Viral drug resistance is a substantial problem given the high rate of viral replication and mutation frequencies. Drug resistant mutants were first recognized for poxviruses with thiosemicarbazone (Appleyard and Way (1966) Brit. J. Exptl. Pathol. 47, 144-51), for poliovirus with guanidine (Melnick et al. (1961) Science 134, 557), for influenza A virus with amantadine (Oxford et al. (1970) Nature 226, 82-83; Cochran et al. (1965) Ann. NY Acad Sci 130, 423-429) and for herpes simplex virus with iododeoxyuridine (Jawetz et al. (1970) Ann. NY Acad Sci 173, 282-291). Approximately 75 HIV drug resistance mutations to various anti-viral agents have been identified to date (Mellors et al. (1995) Intnl. Antiviral News, supplement and Condra, J. H. et al. (1996) J. Virol. 70, 8270-8276).
The small and efficient genomes of viruses have lent themselves to the intensive investigation of the molecular genetics, structure and replicative cycles of most important human viral pathogens. As a consequence, the sites and mechanisms have been characterized for both the activity of and resistance to anti-viral drugs more precisely than have those for any other class of drugs. (Richman (1994) Trends Microbiol. 2, 401-407). The likelihood that resistant mutants will emerge is a function of at least four factors: 1) the viral mutation frequency; 2) the intrinsic mutability of the viral target site with respect to a specific anti-viral; 3) the selective pressure of the anti-viral drug; and, 4) the magnitude and rate of virus replication. With regard to the first factor, for single stranded RNA viruses, whose genome replication lacks a proofreading mechanism, the mutation frequencies are approximately 3×10−5 per base-pair per replicative cycle (Holland et al. (1992) Curr. Topics Microbiol Immunol. 176, 1-20; Mansky et al. (1995) J. Virol. 69, 5087-94; Coffin (1995) Science 267, 483-489). Thus, a single 10 kilobase genome, like that of human immunodeficiency virus (HIV), would be expected to contain on average one mutation for every three progeny viral genomes. As to the second factor, the intrinsic mutability of the viral target site in response to a specific anti-viral agent can significantly affect the likelihood of resistant mutants. For example, zidovudine (AZT) selects for mutations in the reverse transcriptase of HIV more readily in vitro and in vivo than does the other approved thymidine analog d4T (stavudine).
One, perhaps inevitable consequence of the action of an anti-viral drug is that it confers sufficient selective pressure on virus replication to select for drug-resistant mutants (Herrmann et al. (1977) Ann NY Acad Sci 284, 632-7). With respect to the third factor, with increasing drug exposure, the selective pressure on the replicating virus population increases to promote the more rapid emergence of drug resistant mutants. For example, higher doses of AZT tend to select for drug resistant virus more rapidly than do lower doses (Richman et al. (1990) J. AIDS. 3, 743-6). This selective pressure for resistant mutants increases the likelihood of such mutants arising as long as significant levels of virus replication are sustained.
The fourth factor, the magnitude and rate of replication of the virus population, has major consequences on the likelihood of emergence of resistant mutants. Many virus infections are characterized by high levels of virus replication with high rates of virus turnover. This is especially true of chronic infections with HIV as well as hepatitis B and C viruses. The likelihood of emergence of AZT resistance increases in HIV-infected patients with diminishing CD4 lymphocyte counts which are associated with increasing levels of HIV replication (Ibid).
Higher levels of virus increase the probability of preexisting mutants. It has been shown that the emergence of a resistant population results from the survival and selective proliferation of a previously existing subpopulation that randomly emerges in the absence of selective pressure. All viruses have a baseline mutation rate. With calculations of approximately 1010 new virions being generated daily during HIV infection (Ho et al. (1995) Nature 373, 123-126), a mutation rate of 10−4 to 10−5 per nucleotide guarantees the preexistence of almost any mutation at any time point during HIV infection. Evidence is accumulating that drug resistant mutants do in fact exist in subpopulations of HIV infected individuals (Najera et al. (1994) AIDS Res Hum Retroviruses 10, 1479-88; Najera et al. (1995) J. Virol. 69, 23-31). The preexistence of drug resistant picornavirus mutants at a rate of approximately 10−5 is also well documented (Ahmad et al. (1987) Antiviral Res. 8, 27-39).
Human Immunodeficiency Virus (HIV)
Acquired immune deficiency syndrome (AIDS) is a fatal human disease, generally considered to be one of the more serious diseases to ever affect humankind. Globally, the numbers of human immunodeficiency virus (HIV) infected individuals and of AIDS cases increase relentlessly and efforts to curb the course of the pandemic, some believe, are of limited effectiveness. Two types of HIV are now recognized: HIV-1 and HIV-2. By Dec. 31, 1994 a total of 1,025,073 AIDS cases had been reported to the World Health Organization. This is only a portion of the total cases, and WHO estimates that as of late 1994, allowing for underdiagnosis, underreporting and delays in reporting, and based on the estimated number of HIV infections, there have been over 4.5 million cumulative AIDS cases worldwide (Mertens et al. (1995) AIDS 9 (Suppl A), S259-S272). Since HIV began its spread in North America, Europe and sub-Saharan Africa, over 19.5 million men, women and children are estimated to have been infected (Ibid). One of the distinguishing features of the AIDS pandemic has been its global spread within the last 20 years, with about 190 countries reporting AIDS cases today. The projections of HIV infection worldwide by the WHO are staggering. The projected cumulative total of adult AIDS cases by the year 2000 is nearly 10 million. By the year 2000, the cumulative number of HIV-related deaths in adults is predicted to rise to more than 8 million from the current total of around 3 million.
HIV-1 and HIV-2 are enveloped retroviruses with a diploid genome having two identical RNA molecules. The molecular organization of HIV is (5′) U3-R-U5-gag-pol-env-U3-R-U5 (3′) as shown in FIG. 1a. The U3, R, and U5 sequences form the long terminal repeats (LTR) which are the regulatory elements that promote the expression of the viral genes and sometimes nearby cellular genes in infected hosts. The internal regions of the viral RNA code for the structural proteins: gag (p55, p17, p24 and p7 core proteins), pol (p10 protease, p66 and p51 reverse transcriptase and p32 integrase) and env (gp120 and gp41 envelope glycoproteins). Gag codes for a polyprotein precursor that is cleaved by a viral protease into three or four structural proteins; pol codes for reverse transcriptase (RT) and the viral protease and integrase; env codes for the transmembrane and outer glycoprotein of the virus. The gag and pol genes are expressed as a genomic RNA while the env gene is expressed as a spliced subgenomic RNA. In addition to the env gene there are other HIV genes produced by spliced subgenomic RNAs that contribute to the replication and biologic activities of the virus. These genes include: tat which encodes a protein that activates the expression of viral and some cellular genes; rev which encodes a protein that promotes the expression of unspliced or single-spliced viral mRNAs; nef which encodes a myristylated protein that appears to modulate viral production under certain conditions; vif which encodes a protein that affects the ability of virus particles to infect target cells but does not appear to affect viral expression or transmission by cell-to-cell contact; vpr which encodes a virion-associated protein; and vpu which encodes a protein that appears to promote the extracellular release of viral particles.
No disease better exemplifies the problem of viral drug resistance than AIDS. Drug resistant HIV isolates have been identified for nucleoside and non-nucleoside reverse transcriptase inhibitors and for protease inhibitors. The emergence of HIV isolates resistant to AZT is not surprising since AZT and other reverse transcriptase inhibitors only reduce virus replication by about 90%. High rates of virus replication in the presence of the selective pressure of drug treatment provide ideal conditions for the emergence of drug-resistant mutants. Patients at later stages of infection who have higher levels of virus replication develop resistant virus with AZT treatment more quickly than those at early stages of infection (Richman et al. (1990) J AIDS 3, 743-6). The initial description of the emergence of resistance to AZT identified progressive and stepwise reductions in drug susceptibility (Larder et al. (1989) Science 243, 1731-1734). This was explained by the recognition of multiple mutations in the gene for reverse transcriptase that contributed to reduced susceptibility (Larder et al. (1989) Science 246, 1155-1158). These mutations had an additive or even synergistic contribution to the phenotype of reduced susceptibility (Kellam et al. (1992) Proc. Natl. Acad. Sci. 89, 1934-1938). The cumulative acquisition of such mutations resulted in progressive decreases in susceptibility. Similar effects have been seen with non-nucleoside reverse transcriptase inhibitors (Nunberg et al. (1991) J Virol 65, 4887-4892; Sardanna et al. (1992) J Biol Chem 267, 17526-17530). Studies of protease inhibitors have found that the selection of HIV strains with reduced drug susceptibility occurs within weeks (Ho et al. (1994) J Virol 68, 2016-2020; Kaplan et al. (1994) Proc. Natl. Acad. Sci. 91, 5597-5601). While recent studies have shown protease inhibitors to be more powerful than reverse transcriptase inhibitors, nevertheless resistance has developed. (Condra et al., Id. and Report 3rd Conference on Retroviruses and Opportunistic Infections, March 1996). Subtherapeutic drug levels, whether caused by reduced dosing, drug interactions, malabsorption or reduced bioavailability due to other factors, or self-imposed drug holidays, all permit increased viral replication and increased opportunity for mutation and resistance. (Id.)
The selective pressure of drug treatment permits the outgrowth of preexisting mutants. With continuing viral replication in the absence of completely suppressive anti-viral drug activity, the cumulative acquisition of multiple mutations can occur over time, as has been described for AZT and protease inhibitors of HIV. Indeed viral mutants multiply resistant to different drugs have been (observed (Larder et al. (1989) Science 243, 1731-1734; Larder et al. (1989) Science 246, 1155-1158; Condra et al. (1995) Nature 374, 569-71). With the inevitable emergence of resistance in many viral infections, as with HIV for example, strategies must be designed to optimize treatment in the face of resistant virus populations. Ascertaining the contribution of drug resistance to drug failure is a difficult problem because patients who are more likely to develop drug resistance are more likely to have other confounding factors that will predispose them to a poor prognosis (Richman (1994) AIDS Res Hum Retroviruses 10, 901-905). In addition patients contain mixtures of viruses with different susceptibilities.
Hepatitis B (HBV)
HBV is a causative agent for acute and chronic hepatitis, which strikes about 200 million patients worldwide. Zuckerman A. J. Trans. R. Soc. Trop. Med. Hygiene (1982) 76, 711-718. HBV infection acquired in adult life is often clinically inapparent, and most acutely infected adults recover completely from the disease and clear the virus. Rarely, however, the acute liver disease may be so severe that the patient dies of fulminant hepatitis. A small fraction, perhaps 5 to 10%, of acutely infected adults, becomes persistently infected by the virus and develops chronic liver disease of varying severity. Neonatally transmitted HBV infection, however, is rarely cleared, and more than 90% of such children become chronically infected. Because HBV is commonly spread from infected mother to newborn infant in highly populated areas of Africa and Asia, several hundred million people throughout the world are persistently infected by HBV for most of their lives and suffer varying degrees of chronic liver disease, which greatly increases their risk of developing cirrhosis and hepatocellular carcinoma (HCC). Indeed, the risk of HCC is increased 100-fold in patients with chronic hepatitis, and the lifetime risk of HCC in males infected at birth approaches 40%. Beasley R P et al., Lancet (1981) 2, 1129-1133. Accordingly, a large fraction of the world's population suffers from and dies of these late complications of HBV infection. The development of anti-HBV drugs has been long awaited, but has been hampered by the extremely narrow host range of HBV: HBV replicates mainly in human and chimpanzee livers and rot in experimental animals or in cultured cells. Tiollais, P et al. Nature (London) (1985) 317, 489-495.
Hepatitis B virus is a DNA virus with a compact genomic structure; despite its small, circular, 3200 base pairs, HBV DNA codes for four sets of viral products and has a complex, multiparticle structure. HBV achieves its genomic economy by relying on an efficient strategy of encoding proteins from four overlapping genes: S, C, P, and X. HBV is one of a family of animal viruses, hepadnaviruses, and is classified as hepadnavirus type 1. Similar viruses infect certain species of woodchucks, ground and tree squirrels, and Pekin ducks. All hepadnaviruses, including HBV, share the following characteristics: 1) three distinctive morphological forms exist, 2) all members have proteins that are functional and structural counterparts to the envelope and nucleocapsid antigens of HBV, 3) they replicate within the liver but can also exist in extrahepatic sites, 4) they contain an endogenous DNA polymerase with both RNA- and DNA-dependent DNA polymerase activities, 5) their genomes are partially double stranded circular DNA molecules, 6) they are associated with acute and chronic hepatitis and hepatocellular carcinoma and 7) replication of their genome goes through an RNA intermediate which is reverse transcribed into DNA using the virus's endogenous RNA-dependent DNA polymerase activity in a manner analogous to that seen in retroviruses. In the nucleus of infected liver cells, the partially double stranded DNA is converted to a covalently closed circular double stranded DNA (cccDNA) by the DNA-dependent DNA polymerase. Transcription of the viral DNA is accomplished by a host RNA polymerase and gives rise to several RNA transcripts that differ in their initiation sites but all terminate at a common polyadenylation signal. The longest of these RNAs acts as the pregenome for the virus as well as the message for the some of the viral proteins. Viral proteins are translated from the pregenomic RNAs, and the proteins and RNA pregenome are packaged into virions and secreted from the hepatocyte. Although HBV is difficult to cultivate in vitro, several cells have been successfully transfected with HBV DNA resulting in the in vitro production of HBV particles.
There are three particulate forms of HBV: non-infectious 22 nm particles, which appear as either spherical or long filamentous forms, and 42 nm double-shelled spherical particles which represent the intact infectious hepatitis B virion. The envelope protein, HBsAg, is the product of the S gene of HBV and is found on the outer surface of the virion and on the smaller spherical and tubular structures. Upstream of the S gene open reading frame are the pre-S gene open reading frames, pre-S1 and pre-S2, which code for the pre-S gene products, including receptors on the HBV surface for polymerized human serum albumin and the attachment sites for hepatocyte receptors. The intact 42 nm virion can be disrupted by mild detergents and the 27 nm nucleocapsid core particle isolated. The core is composed of two nucleocapsid proteins coded for by the C gene. The C gene has two initiation codons defining a core and a precore region. The major antigen expressed on the surface of the nucleocapsid core is coded for by the core region and is referred to as hepatitis B core antigen (HBcAg). Hepatitis B e antigen (HBeAg) is produced from the same C gene by initiation at the precore ATG.
Also packaged within the nucleocapsid core is a DNA polymerase, which directs replication and repair of HBV DNA. The DNA polymerase is coded for by the P gene, the third and largest of the HBV genes. The enzyme has both DNA-dependent DNA polymerase and RNA-dependent reverse transcriptase activities and is also required for efficient encapsidation of the pregenomic RNA. The fourth gene, X, codes for a small, non-particle-associated protein which has been shown to be capable of transactivating the transcription of both viral and cellular genes.
Although HBV replication is fairly well understood, early steps in HBV infection have not been well defined. Cellular receptors or attachment sites on the virions cannot be studied without appropriate tissue culture assays. In an effort to address this problem, certain cell lines have been developed, human hepatoblastoma cells Huh (HB 611) (Ueda, K. et al., Virology (1989) 169, 213-216) and HepG2 cells (Galle, P. R. and Theilmann, L. Arzneim-Forsch. Drug Res. (1990) 40, 1380-1382) for evaluation of anti-HBV drugs.
Recently, attention has focused on molecular variants of HBV. Variation occurs throughout the HBV genome, and clinical isolates of HBV that do not express viral proteins have been attributed to mutation in individual or even multiple gene locations. For example, variants have been described which lack nucleocapsid proteins, envelope proteins, or both. Two mutants have attracted attention. The first is found in certain patients with severe chronic HBV infection. These patients were found to be infected with an HBV mutant that contained an alteration in the precore region rendering the virus incapable of encoding HBeAg. The most commonly encountered mutation in such patients is a single base substitution, from G to A, which occurs in the second to last codon of the pre-C gene at nucleotide 1896. This substitution results in the replacement of the TGG tryptophan codon by a stop codon (TAG), which prevents the translation of HBeAg. Patients with such precore mutants that are unable to secrete HBeAg tend to have severe liver disease that progresses rapidly to cirrhosis and that does not respond readily to anti-viral therapy.
The second category of HBV mutants consists of escape mutants, in which a single amino acid substitution, from glycine to arginine, occurs at position 145 of the immunodominant a determinant common to all subtypes of HBsAg. This change in HBsAg leads to a critical conformational change that results in a loss of neutralizing activity by anti-HBs antibody.
Presently Available Viral Resistance Assays
The definition of viral drug susceptibility is generally understood to be the concentration of the anti-viral agent at which a given percentage of viral replication is inhibited (e.g. the IC50 for an anti-viral agent is the concentration at which 50% of virus replication is inhibited). Thus, a decrease in viral drug susceptibility is the hallmark that an anti-viral has selected for mutant virus that is resistant to that anti-viral drug. Viral drug resistance is generally defined as a decrease in viral drug susceptibility in a given patient over time. In the clinical context, viral drug resistance is evidenced by the anti-viral drug being less effective or no longer being clinically effective in a patient.
At present the tools available to the researcher and clinician to assess anti-viral drug susceptibility and resistance are inadequate. The classical test for determining the resistance and sensitivity of HIV to an anti-viral agent is complex, time-consuming, expensive, and is hazardous in that it requires the culture of pathogenic virus from each and every patient (Barre-Sinoussi et al (1983) Science 220, 868-871; Popovic et al. (1984) Science 224, 497-500). In this procedure, the patient's peripheral blood mononuclear cells (PBMC) are first cultured to establish a viral stock of known multiplicity of infection (moi), and the viral stock thereby produced is used to infect a target indicator cell line. The resulting burst of viral replication is then typically measured in the presence and absence of an anti-viral agent by determining the production of viral antigens in the cell culture. Such tests can be performed reliably only in the hands of expert investigators, and may take two to three months to carry out at a cost of thousands of dollars per patient for each agent tested. Furthermore, as viral stocks of sufficient moi cannot be established from the PBMC of some HIV patients, the classical test for HIV resistance cannot be performed on all HIV-infected individuals. More significantly, in the course of generating the viral stock by passage of the virus in culture, the characteristics of the viruses themselves can change and may therefore obscure the true nature of the patient's virus. Thus, the application of the classical test has been limited to gathering information about trends in clinical trials and has not been available for a prospective analysis which could be used to custom tailor anti-viral therapy for a given patient. Notwithstanding these limitations, the classical test has two important qualities: it is specific for the agent under evaluation, and it provides information on the phenotype of the patient's own virus, that is, the concentration of the drug which inhibits 50% of viral replication (IC50).
A number of attempts have been made to improve upon the classical test, but each of these has serious shortcomings. The first type of these tests can be described as nonspecific in that they do not determine the characteristics of a patient's own virus at all, but rather provide an independent measure of the course of the infection. Among these tests are those which measure the patient's CD4+ T cell count, the hallmark of HIV disease progression (Goedert et al. (1987) JAMA 257, 331-334), those which measure viral antigen levels (e.g., p24 core antigen (Allain et al. (1987) N. Engl. J. Med. 317, 1114-1121)), and those which measure viral RNA and DNA levels (e.g., quantitative polymerase chain reaction and branched DNA assays (Piatak et al. (1993) Science 259, 1749-1754; Urdea (1993) Clin. Chem. 39, 725-726)). The primary disadvantage of such nonspecific tests is that they do not provide any information on viral drug resistance per se, but rather attempt to infer this information from the apparent course of the patient's disease. In addition, many factors other than viral drug resistance can affect the level of the parameter under consideration. In other words, CD4+ T cell counts, p24 antigen levels and HIV viral RNA levels can vary for reasons other than drug resistance during the course of disease.
Another modified classical test amplifies the viral gene that is the target of the anti-viral agent. In this test the viral gene from a given patient is amplified and then recombined into a biologically active proviral clone of HIV. This proviral clone is transfected into human cells to generate a viral stock of known moi which can then be used to infect a target indicator cell line. In the manner of the classical test, one then determines the production of viral antigens in the presence or absence of the anti-viral agent. One such assay described by Kellam and Larder (1994) Antimicrobial Agents and Chemo. 38, 23-30, involves PCR amplification of reverse transcriptase coding sequences from a patient, which is then introduced into a proviral DNA clone by homologous recombination to reconstitute the complete viral genome including the reverse transcriptase gene which was deleted. The resulting recombinant virus produced from such clones is then cultured in T-cell lines, and the drug sensitivity is tested in the HeLa CD4+ plaque reduction assay. However, this class of test still requires the culturing of virus to determine drug resistance, and is thus difficult, lengthy and costly and requires the laboratory investigator to handle hazardous viral cultures. Furthermore, given the attendant variation of the virus itself during the culture process, the results may be correspondingly inaccurate.
A second class of test attempts to provide specific information on the genotype of the patient's HIV, with the ultimate goal of correlating this genotypic information with the virus' drug resistant phenotype. Indeed, specific amino acid substitutions within viral genes such as reverse transcriptase and protease genes have been shown to correspond to specific levels of viral resistance to reverse transcriptase and protease inhibitors, respectively (Larder et al. (1994) J. Gen Virol. 75, 951-957). A major shortcoming associated with such an analysis is that it is indirect and can be obfuscated by secondary mutations which have been shown to add to or counter the effects of the first mutation. It is the complex interplay of all amino acid residues within a given viral polypeptide which ultimately determines the gene product's activity in the presence or absence of an inhibitor. Thus, a database of vast and impractical proportions would be necessary to interpret the status of drug resistance or sensitivity of a given genotype, given the number of potential amino acid changes in the HIV genome.
A third class of test, a recently developed bacterial-based assay makes use of a molecularly cloned viral gene (specifically, the reverse transcriptase gene) which has been inserted into a bacterial expression vector. Upon transformation of special strains of E. coli which are deficient in the bacterial DNA polymerase I, the cloned reverse transcriptase gene can rescue the growth of the bacteria under selected growth conditions. In making E. coli dependent upon reverse transcriptase for their growth, one can ascertain the effects of certain reverse transcriptase inhibitors on the activity of the viral gene (PCT Application No. WO 95/22622). A major shortcoming with this approach, however, is that the inhibitor may be transported across the cell membrane and metabolized differently by the bacteria than it is by a human cell, and as a result the concentration of the true metabolic inhibitor of the reverse transcriptase may be grossly different in the bacterium than it would be in a relevant human cell target of infection, or the true inhibitor may be absent altogether. Indeed, nucleoside metabolism is known to differ markedly between human and bacterial cells. Another significant shortcoming of this approach is that the assay measures DNA-dependent DNA polymerase activity of reverse transcriptase but not the RNA-dependent DNA polymerase, strand transfer or RNAse H activities of the reverse transcriptase. Thus, an anti-viral compound which acts, at least in part, on these other activities would not have its full inhibitory activity in this assay. Yet another difficulty with this approach is that it is a growth-based test; thus if an inhibitor (eg., a nucleoside analog) also blocks bacterial growth for reasons other than its effects on reverse transcriptase, it can not be adequately tested in this system.
Viral Vectors
Viral vectors and particularly retroviral vectors have been used for modifying mammalian cells because of the high efficiency with which retroviral vectors infect target cells and integrate into the target cell genome. Because of their ability to insert into the genome of mammalian cells much attention has focused on retroviral vectors for use in gene therapy. Details on retroviral vectors and their use can be found in patents and patent publications including European Patent Application EPA 0 178 220, U.S. Pat. No. 4,405,712, PCT Application WO 92/07943, U.S. Pat. No. 4,980,289, U.S. Pat. No. 5,439,809 and PCT Application WO 89/11539. The teachings of these patents and publications are incorporated herein by reference.
One consequence of the emphasis on retroviral vector technology has been the development of packaging cell lines. A major problem with the use of retroviruses is the possibility of the spread of replication-competent retrovirus. There is thus a need for producing helper vectors which could not be processed into virions. As a result packaging defective vectors and packaging cell lines were developed. Details on packaging-defective vectors and packaging cell lines can be found in patents and patent publications, including U.S. Pat. No. 5,124,263, European Patent Application Pub. No. 0386 882, PCT Application No. WO 91/19798, PCT Application No. WO 88/08454, PCT Application No. WO 93/03143, U.S. Pat. No. 4,650,764, U.S. Pat. No. 4,861,719 and U.S. Pat. No. 5,278,056, the disclosures of which are incorporated herein by reference.
It is an object of this invention to provide a drug susceptibility and resistance test capable of showing whether a viral population in a patient is resistant to a given prescribed drug. Another object of this invention is to provide a test that will enable the physician to substitute one or more drugs in a therapeutic regimen for a patient that has become resistant to a given drug or drugs after a previous course of therapy. Yet another object of this invention is to provide a test that will enable selection of an effective drug regimen for the treatment of virus infections. Yet another object of this invention is to provide a safe, standardized, affordable, rapid, precise and reliable assay of drug susceptibility and resistance for clinical and research application. Still another object of this invention is to provide a test and methods for evaluating the biological effectiveness of candidate drug compounds which act on specific viral genes and/or viral proteins particularly with respect to viral drug resistance and cross resistance. It is also an object of this invention to provide the means and compositions for evaluating viral drug resistance and susceptibility. This and other objects of this invention will be apparent from the specification as a whole.
Viral Drug Resistance
The use of anti-viral compounds for chemotherapy and chemoprophylaxis of viral diseases is a relatively new development in the field of infectious diseases, particularly when compared with the more than 50 years of experience with antibacterial antibiotics. The design of anti-viral compounds is not straightforward because viruses present a number of unique problems. Viruses must replicate intracellularly and often employ host cell enzymes, macromolecules, and organelles for the synthesis of virus particles. Therefore, safe and effective anti-viral compounds must be able to discriminate with a high degree of efficiency between cellular and virus-specific functions. In addition, because of the nature of virus replication, evaluation of the in vitro sensitivity of virus isolates to anti-viral compounds must be carried out in a complex culture system consisting of living cells (e.g. tissue culture). The results from such assay systems vary widely according to the type of tissue culture cells which are employed and the conditions of assay.
Viral drug resistance is a substantial problem given the high rate of viral replication and mutation frequencies. Drug resistant mutants were first recognized for poxviruses with thiosemicarbazone (Appleyard and Way (1966) Brit. J. Exptl. Pathol. 47, 144-51), for poliovirus with guanidine (Melnick et al. (1961) Science 134, 557), for influenza A virus with amantadine (Oxford et al. (1970) Nature 226, 82-83; Cochran et al. (1965) Ann. NY Acad Sci 130, 423-429) and for herpes simplex virus with iododeoxyuridine (Jawetz et al. (1970) Ann. NY Acad Sci 173, 282-291). Approximately 140 HIV drug resistance mutations to various anti-viral agents have been identified to date (Mellors et al. (1995) Intnl. Antiviral News, supplement and Condra, J. H. et al. (1996) J. Virol. 70, 8270-8276). Approximately 20 human cytomegalovirus (HCMV) drug resistance mutations to various anti-viral agents have been identified to date (Biron (1996) Antiviral Chemotherapy, 4, 135-143).
The small and efficient genomes of viruses have lent themselves to the intensive investigation of the molecular genetics, structure and replicative cycles of most important human viral pathogens. As a consequence, the sites and mechanisms have been characterized for both the activity of and resistance to anti-viral drugs more precisely than have those for any other class of drugs. (Richman (1994) Trends Microbiol. 2, 401-407). The likelihood that resistant mutants will emerge is a function of at least four factors: 1) the viral mutation frequency; 2) the intrinsic mutability of the viral target site with respect to a specific anti-viral; 3) the selective pressure of the anti-viral drug; and, 4) the magnitude and rate of virus replication. With regard to the first factor, for single stranded RNA viruses, whose genome replication lacks a proofreading mechanism, the mutation frequencies are approximately 3×10−5 per base-pair per replicative cycle (Holland et al. (1992) Curr. Topics Microbiol Immunol. 176, 1-20; Mansky et al. (1995) J Virol. 69, 5087-94; Coffin (1995) Science 267, 483-489). Thus, a single 10 kilobase genome, like that of human immunodeficiency virus (HIV) or hepatitis C virus (HCV), would be expected to contain on average one mutation for every three progeny viral genomes. As to the second factor, the intrinsic mutability of the viral target site in response to a specific anti-viral agent can significantly affect the likelihood of resistant mutants. For example, zidovudine (AZT) selects for mutations in the reverse transcriptase of HIV more readily in vitro and in vivo than does the other approved thymidine analog d4T (stavudine).
One, perhaps inevitable consequence of the action of an anti-viral drug is that it confers sufficient selective pressure on virus replication to select for drug-resistant mutants (Herrmann et al. (1977) Ann NY Acad Sci 284, 632-7). With respect to the third factor, with increasing drug exposure, the selective pressure on the replicating virus population increases to promote the more rapid emergence of drug resistant mutants. For example, higher doses of AZT tend to select for drug resistant virus more rapidly than do lower doses (Richman et al. (1990) J. AIDS. 3, 743-6). This selective pressure for resistant mutants increases the likelihood of such mutants arising as long as significant levels of virus replication are sustained.
The fourth factor, the magnitude and rate of replication of the virus population, has major consequences on the likelihood of emergence of resistant mutants. Many virus infections are characterized by high levels of virus replication with high rates of virus turnover. (Perelson et al. (1996) Science, 271, 1582-1586; Nowak et al. (1996), PNAS 93, 4398-4402). This is especially true of chronic infections with HIV as well as hepatitis B and C viruses. The likelihood of emergence of AZT resistance increases in HIV-infected patients with diminishing CD4 lymphocyte counts which are associated with increasing levels of HIV replication (Ibid).
Higher levels of virus increase the probability of preexisting mutants. It has been shown that the emergence of a resistant population results from the survival and selective proliferation of a previously existing subpopulation that randomly emerges in the absence of selective pressure. All viruses have a baseline mutation rate. With calculations of approximately 1010 new virions being generated daily during HIV infection (Ho et al. (1995) Nature 373, 123-126), a mutation rate of 10−4 to 10−5 per nucleotide guarantees the preexistence of almost any single point mutation at any time point during HIV infection. Evidence is accumulating that drug resistant mutants do in fact exist in subpopulations of HIV infected individuals (Najera et al. (1994) AIDS Res Hum Retroviruses 10, 1479-88; Najera et al. (1995) J. Virol. 69, 23-31; Havlir et al. (1996) J. Virol., 70, 7894-7899). The preexistence of drug resistant picornavirus mutants at a rate of approximately 10−5 is also well documented (Ahmad et al. (1987) Antiviral Res. 8, 27-39).
Hepatitis C Virus (HCV)
Hepatitis C virus (HCV) infection occurs throughout the world and, prior to its identification, represented the major cause of transfusion-associated hepatitis. The seroprevalence of anti-HCV in blood donors from around the world has been shown to vary between 0.02% and 1.23%. HCV is also a common cause of hepatitis in individuals exposed to blood products. There have been an estimated 150,000 new cases of HCV infection each year in the United States alone during the past decade (Alter 1993, Infect. Agents Dis. 2, 155-166; Houghton 1996, in Fields Virology, 3rd Edition, pp. 1035-1058).
The hepatitis C virus (HCV) is a member of the flaviviridae family of viruses, which are positive stranded, non-segmented, RNA viruses with a lipid envelope. Other members of the family are the pestiviruses (e.g. bovine viral diarrheal virus, or BVDV, and classical swine fever virus, or CSFV), and flaviviruses (e.g. yellow fever virus and Dengue virus). See Rice, 1996 in Fields Virology, 3rd Edition, pp. 931-959. Molecular dissection of HCV replication and hence understanding the functions of its encoded proteins, while greatly advanced by the isolation of the virus and sequencing of the viral genome, has been hampered by the lack of an efficient cell culture system for production of native or recombinant HCV from molecular clones. However, low-level replication has been observed in several cell lines infected with virus from HCV-infected humans or chimpanzees, or transfected with RNA derived from cDNA clones of HCV.
HCV replicates in infected cells in the cytoplasm, in close association with the endoplasmic reticulum (see FIG. 1). Incoming positive sense RNA is released and translation is initiated via an internal initiation mechanism (Wang et al. 1993, J. Virol. 67, 3338-3344; Tsukiyama-Kohara et al. 1992, J. Virol. 66, 1476-1483). Internal initiation is directed by a cis-acting RNA element at the 5′ end of the genome; some reports have suggested that full activity of this internal ribosome entry site, or IRES, is seen with the first 700 nucleotides, which spans the 5′ untranslated region (UTR) and the first 123 amino acids of the open reading frame (ORF) (Lu and Wimmer, PNAS 93, 1412, 1996). All of the protein products of HCV are produced by proteolytic cleavage of a large (3010-3030 amino acids, depending on the isolate) polyprotein, carried out by one of three proteases: the host signal peptidase, the viral self-cleaving metalloproteinase, NS2, or the viral serine protease NS3/4A (see FIG. 2). The combined action of these enzymes produces the structural proteins (C, E1 and E2) and non-structural (NS2, NS3, NS4A, NS4B, NS5A, and NS5B) proteins which are required for replication and packaging of viral genomic RNA. NS5B is the viral RNA-dependent RNA polymerase (RDRP) that is responsible for the conversion of the input genomic RNA into a negative stranded copy (complimentary RNA, or cRNA); the cRNA then serves as a template for transcription by NS5B of more positive sense genomic/messenger RNA.
Several institutions and laboratories are attempting to identify and develop anti-HCV drugs. Currently the only effective therapy against HCV is alpha-interferon, which can control the amount of virus in the liver and blood (viral load) in only a small proportion of infected patients (Houghton 1996, in Fields Virology, 3rd Edition, pp. 1035-1058). However, given the availability of the molecular structure of the HCV serine protease, NS3/4A (Love et al., 1996, Cell 87, 331-342; Kim et al. 1996, Cell 87, 343-355), and success using protease inhibitors in the treatment of HIV-1 infection, there should soon be alternatives available. In addition to HCV protease inhibitors, other inhibitors which might specifically interfere with HCV replication could target virus specific activities such as internal initiation directed by the IRES, RDRP activity encoded by NS5B, or RNA helicase activity encoded by NS3.
As a result of a high error rate of their RDRPs, RNA viruses are particularly able to adapt to many new growth conditions. Most polymerases in this class have an estimated error rate of 1 in 10,000 nucleotides copied. With a genome size of approximately 9.5 kb, at least one nucleotide position in the genome of HCV is likely to sustain a mutation every time the genome is copied. It is therefore likely for drug resistance to develop during chronic exposure to an anti-viral agent. As in the case of HIV, a rapid and convenient assay for drug resistant HCV would greatly improve the likelihood of successful antiviral therapy, given a selection of drugs and non-overlapping patterns of drug resistant genotypes. Resistance-associated mutations can sometimes be identified rapidly by growing the virus in cell culture in the presence of the drug, an approach used with considerable success for HIV-1. To date, however, a convenient cell culture system for HCV is lacking. It is therefore not possible to determine the precise nature of genetic changes which confer drug resistance in vitro. Thus, in the absence of a list of known resistance-associated mutations, the preferred resistance assay is one that relies on a phenotypic readout rather than a genotypic one.
Presently Available Drug Resistance Assays
There are no well established drug resistance assays for HCV currently available. However, several investigators have devised chimeric virus systems containing the HCV NS3 protease such that replication of the chimera, or expression of a reporter gene, is dependent on NS3 activity. These systems were devised in order to study various aspects of NS3 function and its cofactor, NS4A, and to serve as prototypes for cell-based drug screening assays.
Hirowatari et al (Anal. Biochem. 225:113, 1995) constructed an expression vector to synthesize an endoplasmic reticulum-tethered NS2-NS3-tax1 fusion protein (tax1 is the transcriptional trans-activator from the HTLV-I retrovirus). Upon cleavage by NS3 (or NS2), the tax1 transactivator is released and migrates to the nucleus, where it acts to activate the expression of a reporter gene (CAT) controlled by the HTLV-1 LTR.
Hahm et al (Virology 226:318, 1996) constructed a chimeric poliovirus containing HCV NS3 upstream of the poliovirus polyprotein; replication of the chimera is dependent on NS3 protease activity, since liberation of the native N-terminus of the poliovirus polyprotein is essential for initiation of poliovirus replication. However, this chimera does not include the NS4A protein of HCV, which has been shown to modify the activity of NS3.
Filocamo et al (J. Virol. 71:1417, 1997) constructed a chimeric Sindbis virus containing HCV NS3/4A upstream of the Sindbis virus polyprotein; replication of the chimera is dependent on NS3 protease activity, since liberation of the native N-terminus of the Sindbis virus polyprotein is essential for initiation of Sindbis virus replication.
In Vitro Expression Systems
Resistance test vectors rely on a cell culture system for transfection, replication, and expression of the indicator gene. Others have described systems for transfection of intact HCV RNA synthesized in vitro from cDNA constructs into Huh7 cells, and have demonstrated that replication can occur (Yoo et al. (1995) J. Virol. 69, 32-38). In addition, several cell lines have been identified which support HCV replication following infection with virus present in HCV-infected human or chimpanzee serum or plasma (Valli et al. (1997) Res. Virol. 148, 181-186; Shimizu et al. (1992) PNAS 89, 5477-5481; Shimizu et al. (1993) PNAS 90, 6037-6041; Shimizu and Yoshikura (1994) J. Virol. 68, 8406-8408; Shimizu et al. (1994) J. Virol. 68, 1494-1500; Mizutani et al. (1996) J. Virol. 70, 7219-7223). Currently these systems are limited to those in which replication is detected by sensitive methods such as RT/PCR, and are unlikely to allow for efficient production of an indicator gene (IG). Improved methods may soon become available as new cell lines and transfection methods are discovered. One example of a potential improvement would be to transfect the RNA as RNP complexes, prepared by in vitro transcription in the presence of purified NS5B, so that transcription can commence immediately upon uptake into the cells; this strategy has been applied to the negative-stranded RNA viruses such as influenza virus (Enami and Palese (1991) J. Virol. 65, 2711-2713, rabies virus (Schnell et al. (1994) EMBO J. 13, 4195-4203), and vesicular stomatitis virus (Lawson et al. (1995) PNAS 92, 4477-4481).
Since the genome length RNA of flaviviruses is infectious, HCV vectors may be in the form of a cDNA construct containing a promoter for the T7 RNA polymerase at the 5 end, and a T7 polymerase terminator sequence at the 3 end. Thus RNA can be synthesized in large quantities in vitro and transfected into cells. An alternative approach is to transfect DNA constructs, which contain a strong eucaryotic promoter (such as the CMV IE promoter), directly into cells. Potential advantages of transfection of RNA, rather than DNA, include the following: transfection of RNA circumvents potential cell-type specific restrictions of promoter activity; translation of recombinant protein usually occurs within minutes to hours following transfection of biologically active RNA, whereas translation following DNA transfection must be preceded by transcription and RNA processing events, which incurs delays of hours to days before maximal expression levels are reached. Representation of viral quasispecies is more straightforward when transfecting RNA since sufficient quantities of RNA can be synthesized from uncloned PCR products by including the sequence of a bacteriophage RNA polymerase in the 5′ PCR primer. This approach has been shown to preserve quasispecies diversity of poliovirus (Chumakov, J. Virol. 70:7331-7334, 1996). Generation of precise 5′ and 3′ termini on the RNA is more easily achieved by in vitro transcription, through the placement of the promoter sequence at the 5′ end and a restriction endonuclease recognition site at the 3′ end (used for DNA template linearization prior to transcription) relative to the viral sequences. However, the 3′ terminus may also be generated precisely via the placement of a self-cleaving RNA ribozyme sequence, present in a cDNA construct (e.g. see Chowrira et al. (1994), J. Biol. Chem. 269, 25856-25864.).
A third transfection strategy, which possesses some of the advantages of RNA transfection, is DNA transfection of constructs containing a T7 RNA polymerase promoter at the 5′ end, and a T7 RNA polymerase terminator at the 3′ end, into cells which express the T7 RNA polymerase. Expression of the polymerase may be achieved by various means, perhaps the most efficient of these being the infection of the transfected cells with a recombinant T7 polymerase/vaccinia virus (Fuerst et al. (1986) PNAS 83, 8122-8126.)
Human Cytomegalovirus (HCMV)
Human Cytomegalovirus (HCMV) is endemic throughout the world and infection rates appear to be relatively constant throughout the year rather than seasonal. Humans are the only known reservoir for HCMV and natural transmission occurs by direct or indirect person-to-person contact. Between 0.2% and 2.2% of infants born in the United States are infected in utero. Another 8% to 60% become infected during the first six months of life as a result of infection acquired during birth or following breast feeding. Because of the high incidence of reactivation of HCMV infection in the breast, breast milk transmission could represent the most common mode of HCMV transmission worldwide. In most developed countries, 40% to 80% of children are infected before puberty. In other areas of the world, 90% to 100% of the population become infected during childhood.
Human cytomegalovirus (HCMV) is a member of the herpesvirus family. A typical herpes virion consists of a core containing a linear double-stranded DNA and icosadeltahedral capsid approx. 100-110 nm in diameter containing 162 capsomeres with a hole running down the long axis, an amorphous “tegument” that surrounds the core and an envelope containing viral glycoprotein spikes on its surface. Virion sizes range from 120-300 nm due to differences in the thickness of the tegument layer. There are three subgroups of herpesviruses:
1. Alphaherpesvirinae: HSV, VZV. variable host range, relatively short reproductive cycle, rapid spread in culture, efficient destruction of infected cells, capacity to establish latent infections in sensory ganglia.
2. Betaherpesvirinae: HCMV. Restricted host range, long reproductive cycle, slow progression of infection in culture. Infected cells become enlarged and carrier cells are readily established. Virus can be maintained in latent form in secretory glands, lymphoreticular cells, kidneys and other tissues.
3. Gammaherpesvirinae: EBV. experimental host range extremely narrow, replicate in lymphoblastoid cells and cause lytic infections in some types of epithelial and fibroblastoid cells.
There are 8 known human herpesviruses: Human herpesvirus 1 (Herpes simplex virus 1, HSV-1), Human herpesvirus 2 (Herpes simplex virus 2, HSV-2), Human herpesvirus 3 (Varicella-zoster virus, VZV), Human herpesvirus 4 (Epstein-Barr virus, EBV), Human herpesvirus 5 (Human cytomegalovirus), Human herpesvirus 6, Human herpesvirus 7, and Human herpesvirus 8. The genomes of herpes viruses consist of a linear double-stranded (ds) DNA in the virion which circularizes and concatamerizes upon release from the virus capsid in the nucleus of infected cells (See FIG. 10). The genomes of herpesviruses range in size from 120 to 230 kilobase pairs (kbp). The genomes are polymorphic in size (up to 10 kbp differences) within an individual population of virus. This variation is due to the presence of terminal and internal reiterated sequences. Herpes viruses can be classified into six groups, A through F, based on their overall genome organization. HSV and HCMV fall into group E, in which sequences from both termini are repeated in an inverted orientation and juxtaposed internally, dividing the genomes into two components, L (long) and S (short), each of which consists of unique sequences, UL and US, flanked by inverted repeats (FIG. 10). In these viruses both components can invert relative to each other and DNA extracted form virions consists of four equimolar populations differing in the relative orientation of the two components (See FIG. 11).
HCMV is a betaherpesvirus and is unique among the betaherpesvirinae in that it falls into the class E genome type. The genome of HCMV is approximately 230 kbp in length and has been completely sequenced (EMBL Seq database accession # X17403). In a naturally occurring population of virus the genome exists in 4 isomers (See FIG. 11). In HCMV, as in HSV, the L-S junction can be deleted, thereby freezing the genome in one of four isomers without dramatically affecting the ability of the virus to grow in cultured cells.
The HCMV genome contains terminal repeat sequences “a” and “a′” present in a variable number in direct orientation at both ends of the linear genome. A variable number of “a” repeats are also present in an inverted orientation at the L-S junction. The number of “a” sequences in these locations ranges from 1-10 with 1 predominating. The size of “a” in HCMV ranges from 700-900 bp. The “a” sequence carries the cleavage and packaging signal. The packaging signals are two highly conserved short sequence elements located within “a” designated pac-1 and pac-2. A 220-bp fragment that carries both the pac-1 and pac-2 elements is sufficient to convey sites for cleavage/packaging as well as inversion on a recombinant CMV construct. The termini of the linear genome are generated by a cleavage event that leaves a single 3′ overhanging nucleotide at either end of the genome. The genome is further characterized by large inverted repeats called “b” and “b′” (or TRL and IRL) and “c” and “c′” (or IRS and TRS) that flank unique sequences UL and US that make up the L and S components of the genome (See FIG. 10)
The HCMV replication cycle is relatively slow compared to other herpesviruses. Viral replication involves the ordered expression of consecutive sets of viral genes. These sets are expressed at different times after infection and include the α (immediate early), β1 and β2 (delayed early), and γ1 and γ2 (late) sets based on the time after infection that their transcripts accumulate. DNA replication, genome maturation and virion morphogenesis are coordinated through the temporal regulation of the appropriate gene products required for each step. Expression of a gene products is rapid. Late gene expression is delayed for 24-36 hours. Progeny virions begin to accumulate 48 hours post-infection and reach maximal levels at 72-96 hours. In permissive fibroblasts, DNA replication can be detected as early as 14-16 hours post-infection. HCMV stimulates host DNA, RNA and protein synthesis. HCMV replicates more rapidly in actively dividing cells and HCMV replication is inhibited by pretreating cells with agents that reduce host cell metabolism. The HCMV genome circularizes soon after infection. Circles give rise to concatamers and genomic inversion occurs within concatameric forms of the DNA. The majority of replicating DNA is larger than unit length and lacks terminal fragments based on southern blot analysis.
Targets for Drug Resistance
The drugs currently used to treat HCMV (ganciclovir (GCV), foscarnet, cidofovir) are known to select for mutations in two viral genes, the UL97 phosphotransferase and the UL54 viral DNA polymerase.
UL97: phosphotransferase 707 amino acids (aa) (2121 bp). Mutations associated with GCV resistance include aa#: 460, 520, 590, 591, 592, 593, 594, 595, 596, 600, 603, 607, 659, 665. The phosphotransferase protein has two functional domains, 1) the amino terminal 300 aa code for a regulatory domain and 2) the carboxy terminal 400 aa define the catalytic domain. All known drug-resistance mutations are found in the catalytic domain (approx 1.2 kb of sequence). In HSV the thymidine kinase gene product (TK) is responsible for the phosphorylation of GCV in cells and resistance to GCV in HSV is associated with mutations in the thymidine kinase gene. HCMV has no homolog to the HSV thymidine kinase gene. The gene homologous to UL97 in HSV (UL13) is a protein kinase.
UL54: viral DNA polymerase, 1242 a.a. (3726 bp). Mutations in this gene can result in resistance to GCV and other nucleoside analogs (including cidofovir) as well as foscarnet. Mutations associated with foscarnet resistance include aa #: 700 and 715. Mutations associated with GCV resistance include aa#: 301, 412, 501, 503, and 987. The mature protein has four recognized domains: 1) a 5′-3′ exoRNase H, a 3′-5′ exonuclease, a proposed catalytic domain and an accessory protein binding domain.
New therapies in development include agents targeted to the CMV protease (UL80) and the DNA maturational enzyme (“terminase”).
GCV-resistant HCMV has been recovered from the central nervous system (CNS) of patients with HCMV-associated neurologic disease who had received long-term GCV maintenance therapy. Resistant strains of HCMV may be selected and preferentially located in the CNS. It is frequently not possible to culture virus from the cerebral spinal fluid (CSF) but it is possible to amplify HCMV DNA using PCR.
Primary isolates of CMV may replicate slowly. In addition, there is a marked delay in the growth rate of some of the drug resistant clinical isolates. In a mixed virus population, a resistant virus population could be masked by a sensitive one. Thus assay results that depend on the growth of virus could be unreliable.
Most assays for viral culture use blood or urine, because they are easy to obtain. However, the virus from these compartments may not represent the virus in specific tissues where disease is occurring (especially vitreous fluid and Csf). Although there are a few amino acid residues that are modified relatively frequently among drug-resistant strains of herpesviruses recovered from patients, the broad distribution of mutations in the majority of strains makes rapid genetic screening methods impractical. Importantly, since the drug-susceptibility phenotypes resulting from individual genetic changes are complex and variable, a biological test for anti-viral susceptibility of HCMV would be more informative.
Presently Available Viral Resistance Assays
The definition of viral drug susceptibility is generally understood to be the concentration of the anti-viral agent at which a given percentage of viral replication is inhibited (e.g. the IC50 for an anti-viral agent is the concentration at which 50% of virus replication is inhibited). Thus, a decrease in viral drug susceptibility is the hallmark that an anti-viral has selected for mutant virus that is resistant to that anti-viral drug. Viral drug resistance is generally defined as a decrease in viral drug susceptibility in a given patient over time. In the clinical context, viral drug resistance is evidenced by the anti-viral drug being less effective or no longer being clinically effective in a patient.
Several types of assays are available to detect and measure antiviral drug susceptibility of HCMV. The two most commonly used methods are a plaque reduction assay and a DNA hybridization assay. At present the plaque reduction assay is considered the standard. Both assays require HCMV isolation and passage in cell culture. Generally, it takes four to six weeks to obtain the results from the assays.
Plaque reduction assays with increased sensitivity can now be performed directly on clinical specimens, including blood, urine, bronchoalveolar lavage, and cerebrospinal fluid. Two assays which are modified from the standard plaque reduction assay detect either the CMV immediate-early antigen or late antigen. The procedure is essentially the same as the standard plaque reduction assay except that the virus is tested directly without prior passage and the incubation time is reduced to ninety-six hours (Gerna et al. (1995) J. Clin. Microbiol. 33, 738-741). The limitation of these assays is that they can only be performed in patients with high level of viremia. Virus culture remains an essential step in the detection of drug resistant isolates.
An alternative approach is the detection of specific viral DNA mutations related to drug resistance. In this assay, PCR primers are used to amplify viral DNA and restriction sites present in mutant viral DNA but not wildtype DNA are used to determine the genotype of the viral DNA. It is suggested that the analysis of two PCR products with a total of three or four restriction digests is adequate to detect 78-83% of UL97 (certain mutations of UL97 which codes for a phosphotransferase, result in resistance to ganciclovir) mutants resistant to ganciclovir (Chou et al. (1995) J. Infect. Dis. 172, 239-242.). The main limitation of this assay is that infrequent or new resistance mutations are not identified. Also, DNA polymerase mutations (UL54) which are indicative of high-level ganciclovir resistance and a high probability of multidrug resistance are not detected.
It is an object of this invention to provide a drug susceptibility and resistance test capable of showing whether a viral population in a patient is resistant to a given prescribed drug. Another object of this invention is to provide a test that will enable the physician to substitute one or more drugs in a therapeutic regimen for a patient that has become resistant to a given drug or drugs after a previous course of therapy. Yet another object of this invention is to provide a test that will enable selection of an effective drug regimen for the treatment of virus infections. Yet another object of this invention is to provide a safe, standardized, affordable, rapid, precise and reliable assay of drug susceptibility and resistance for clinical and research application. Still another object of this invention is to provide a test and methods for evaluating the biological effectiveness of candidate drug compounds which act on specific viral genes and/or viral proteins particularly with respect to viral drug resistance and cross resistance. It is also an object of this invention to provide the means and compositions for evaluating viral drug resistance and susceptibility. This and other objects of this invention will be apparent from the specification as a whole.